Background: In Western countries, increasing maternal age has led to more pregnancies with a child with Down syndrome (DS). However, prenatal screening programs, diagnostic testing and termination of pregnancy influence the actual DS live birth (LB) prevalence as well. The aim of this study is to examine these factors in the Netherlands for the period 1991–2015. In our study, we establish a baseline for DS LB prevalence before non-invasive prenatal testing will be made available to all pregnant women in the Netherlands in 2017. Methods: Full nationwide data from the Dutch cytogenetic laboratories were used to evaluate the actual DS LB prevalence. In addition, nonselective DS prevalence, which is the DS LB prevalence that would be expected in absence of termination of pregnancies, was estimated on the basis of maternal age distribution in the general population. Results: Because of an increase in maternal age, nonselective DS prevalence increased from around 15.6 [95% confidence interval (CI) 13.9–17.4] per 10 000 LBs in 1991 (311 children in total) to around 22.6 (95% CI 20.3–24.9) per 10 000 in 2015 (385), the increase levelling off in recent years. Actual LB prevalence rose from around 11.6 (95% CI 10.9–12.2) per 10 000 in 1991 (230 children) to an estimated peak of 15.9 (95% CI 15.6–16.2) per 10 000 in 2002 (322), gradually decreasing since to 11.1 (95% CI 10.8–11.5) per 10 000 in 2015 (190). Reduction of DS LBs resulting from elective terminations had been fairly constant between 1995 and 2002 at around 28% and rose afterwards from 35% in 2003 to around 50% in 2015. Conclusions: In spite of expansion of antenatal screening in the Netherlands in the 1990s and early 2000s, actual DS LB prevalence increased during this period. However, after 2002, this trend reversed, probably because of informing all pregnant women about prenatal testing since 2004 and the implementation of a national screening program in 2007.